In recent years, as computer technique and IP network have developed, push service is widely provided in the Internet or intranet. The push service transmits data of a host apparatus in the network to a low-order apparatus of a user or the like at timing different from that of a request from the low-order apparatus. To transmit the data from the host apparatus to the low-order apparatus when the low-order apparatus requests the data from the host apparatus is called pull. In this push service, it is important that data are distributed efficiently and quickly, and realization of the push service which distributes data efficiently and quickly is realized.
Most of the conventional push services are realized by automatically accessing from a user terminal to a predetermined type of information per constant time. That is, a function for accessing to a data source including the predetermined information per constant time is provided to the user terminal, and update information of the data source is obtained from the accessed results.
In another push services, user terminals are registered in a data source including various data, and when the data source is updated, the data source directly provides the updated data to the respective user terminals. In this case, the data source itself has the function for providing distribution service.
However, the conventional push services where the access is executed per constant time, since a check is made as to whether or not the information in the data source is updated in each access, there arises a problem that useless traffic on the network increase.
In addition, in this push service system, since free time intervals between the checks of the updated information in the data source exist, there arises a problem that emergent information such as news flash cannot be obtained immediately.
Meanwhile, in the conventional push service where the user terminals are registered in the respective data sources, since information about the user terminals is registered in the respective data sources, the respective data sources hold the information about the reduplicate user terminals. As a result, a lot of resources are required so that the resources cannot be used effectively, and flexible expandability to the push service system cannot be expected.